Face Generation¶

In this project, you'll use generative adversarial networks to generate new images of faces.

Get the Data¶

You'll be using two datasets in this project:

MNIST
CelebA

Since the celebA dataset is complex and you're doing GANs in a project for the first time, we want you to test your neural network on MNIST before CelebA. Running the GANs on MNIST will allow you to see how well your model trains sooner.

If you're using FloydHub, set data_dir to "/input" and use the FloydHub data ID "R5KrjnANiKVhLWAkpXhNBe".

data_dir = './data'

# FloydHub - Use with data ID "R5KrjnANiKVhLWAkpXhNBe"
#data_dir = '/input'


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import helper

helper.download_extract('mnist', data_dir)
helper.download_extract('celeba', data_dir)

Found mnist Data
Found celeba Data

Explore the Data¶

MNIST¶

As you're aware, the MNIST dataset contains images of handwritten digits. You can view the first number of examples by changing show_n_images.

show_n_images = 25

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
%matplotlib inline
import os
from glob import glob
from matplotlib import pyplot

mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'mnist/*.jpg'))[:show_n_images], 28, 28, 'L')
pyplot.imshow(helper.images_square_grid(mnist_images, 'L'), cmap='gray')

<matplotlib.image.AxesImage at 0x7f18f14be518>

CelebA¶

The CelebFaces Attributes Dataset (CelebA) dataset contains over 200,000 celebrity images with annotations. Since you're going to be generating faces, you won't need the annotations. You can view the first number of examples by changing show_n_images.

show_n_images = 25

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'img_align_celeba/*.jpg'))[:show_n_images], 28, 28, 'RGB')
pyplot.imshow(helper.images_square_grid(mnist_images, 'RGB'))

<matplotlib.image.AxesImage at 0x7f18f13a15f8>

Preprocess the Data¶

Since the project's main focus is on building the GANs, we'll preprocess the data for you. The values of the MNIST and CelebA dataset will be in the range of -0.5 to 0.5 of 28x28 dimensional images. The CelebA images will be cropped to remove parts of the image that don't include a face, then resized down to 28x28.

The MNIST images are black and white images with a single color channel while the CelebA images have 3 color channels (RGB color channel).

Build the Neural Network¶

You'll build the components necessary to build a GANs by implementing the following functions below:

model_inputs
discriminator
generator
model_loss
model_opt
train

Check the Version of TensorFlow and Access to GPU¶

This will check to make sure you have the correct version of TensorFlow and access to a GPU

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
from distutils.version import LooseVersion
import warnings
import tensorflow as tf

# Check TensorFlow Version
assert LooseVersion(tf.__version__) >= LooseVersion('1.0'), 'Please use TensorFlow version 1.0 or newer.  You are using {}'.format(tf.__version__)
print('TensorFlow Version: {}'.format(tf.__version__))

# Check for a GPU
if not tf.test.gpu_device_name():
    warnings.warn('No GPU found. Please use a GPU to train your neural network.')
else:
    print('Default GPU Device: {}'.format(tf.test.gpu_device_name()))

TensorFlow Version: 1.1.0
Default GPU Device: /gpu:0

Input¶

Implement the model_inputs function to create TF Placeholders for the Neural Network. It should create the following placeholders:

Real input images placeholder with rank 4 using image_width, image_height, and image_channels.
Z input placeholder with rank 2 using z_dim.
Learning rate placeholder with rank 0.

Return the placeholders in the following the tuple (tensor of real input images, tensor of z data)

import problem_unittests as tests

def model_inputs(image_width, image_height, image_channels, z_dim):
    """
    Create the model inputs
    :param image_width: The input image width
    :param image_height: The input image height
    :param image_channels: The number of image channels
    :param z_dim: The dimension of Z
    :return: Tuple of (tensor of real input images, tensor of z data, learning rate)
    """
    # TODO: Implement Function
    input_real = tf.placeholder(tf.float32, 
                                shape=(None, image_width, image_height, image_channels),
                                name='input_real')
    inputs_z = tf.placeholder(tf.float32, (None, z_dim), name='input_z')
    learning_rate = tf.placeholder(tf.float32, name='learning_rate')

    return input_real, inputs_z, learning_rate


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_inputs(model_inputs)

Tests Passed

Discriminator¶

Implement discriminator to create a discriminator neural network that discriminates on images. This function should be able to reuse the variables in the neural network. Use tf.variable_scope with a scope name of "discriminator" to allow the variables to be reused. The function should return a tuple of (tensor output of the discriminator, tensor logits of the discriminator).

def leaky_relu(x, alpha=0.2):
    return tf.maximum(alpha*x,x)

def discriminator(images, reuse=False):
    """
    Create the discriminator network
    :param images: Tensor of input image(s)
    :param reuse: Boolean if the weights should be reused
    :return: Tuple of (tensor output of the discriminator, tensor logits of the discriminator)
    """
    # TODO: Implement Function
    with tf.variable_scope('discriminator', reuse=reuse):
        conv_layer1 = tf.layers.conv2d(images, 64, 5, strides=2, padding='same')
        relu1 = leaky_relu(conv_layer1)
        
        conv_layer2 = tf.layers.conv2d(conv_layer1, 128, 5, strides=2, padding='same')
        conv_layer2 = tf.layers.batch_normalization(conv_layer2, training=True)
        relu2 = leaky_relu(conv_layer2)
        
        conv_layer3 = tf.layers.conv2d(conv_layer2, 256, 5, strides=2, padding='same')
        conv_layer3 = tf.layers.batch_normalization(conv_layer3, training=True)
        relu3 = leaky_relu(conv_layer3)
        
        conv_layer4 = tf.layers.conv2d(conv_layer3, 512, 5, strides=2, padding='same')
        conv_layer4 = tf.layers.batch_normalization(conv_layer4)
        relu4 = leaky_relu(conv_layer4)
        
        #Flat
        flat = tf.reshape(relu3, (-1, 4*4*256))
        
        #Logits
        logits = tf.layers.dense(flat, 1)
        out = tf.sigmoid(logits)

    return out, logits


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_discriminator(discriminator, tf)

Tests Passed

Generator¶

Implement generator to generate an image using z. This function should be able to reuse the variables in the neural network. Use tf.variable_scope with a scope name of "generator" to allow the variables to be reused. The function should return the generated 28 x 28 x out_channel_dim images.

def generator(z, out_channel_dim, is_train=True):
    """
    Create the generator network
    :param z: Input z
    :param out_channel_dim: The number of channels in the output image
    :param is_train: Boolean if generator is being used for training
    :return: The tensor output of the generator
    """
    # TODO: Implement Function
    with tf.variable_scope('generator', reuse=False if is_train==True else True):
        #Fully connected layer
        fully_conn = tf.layers.dense(z, 7*7*512)
        fully_conn = tf.reshape(fully_conn, (-1,7,7,512))
        fully_conn = tf.layers.batch_normalization(fully_conn, training=is_train)
        fully_conn = leaky_relu(fully_conn)
        # 7x7x512

        #Convolutions
        conv_layer1 = tf.layers.conv2d_transpose(fully_conn, 256, 3, strides=2, padding='same')
        conv_layer1 = tf.layers.batch_normalization(conv_layer1, training=is_train)
        conv_layer1 = leaky_relu(conv_layer1)
        # 14x14x256

        conv_layer2 = tf.layers.conv2d_transpose(conv_layer1, 128, 3, strides=1, padding='same')
        conv_layer2 = tf.layers.batch_normalization(conv_layer2, training=is_train)
        conv_layer2 = leaky_relu(conv_layer2)
        # 14x14x128

        conv_layer3 = tf.layers.conv2d_transpose(conv_layer2, 64, 3, strides=2, padding='same')
        conv_layer3 = tf.layers.batch_normalization(conv_layer3, training=is_train)
        conv_layer3 = leaky_relu(conv_layer3)
        # 28x28x64   

        # Output layer
        logits = tf.layers.conv2d_transpose(conv_layer3, out_channel_dim, 3, strides=1, padding='same')
        # 28x28xout_dim 

        gen_out = tf.tanh(logits)

    return gen_out


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_generator(generator, tf)

Tests Passed

Loss¶

Implement model_loss to build the GANs for training and calculate the loss. The function should return a tuple of (discriminator loss, generator loss). Use the following functions you implemented:

discriminator(images, reuse=False)
generator(z, out_channel_dim, is_train=True)

def model_loss(input_real, input_z, out_channel_dim):
    """
    Get the loss for the discriminator and generator
    :param input_real: Images from the real dataset
    :param input_z: Z input
    :param out_channel_dim: The number of channels in the output image
    :return: A tuple of (discriminator loss, generator loss)
    """
    # TODO: Implement Function
    # Label smoothing 
    smooth = 0.1
    
    g_model = generator(input_z, out_channel_dim, is_train=True)
    d_model_real, d_logits_real = discriminator(input_real, reuse=False)
    d_model_fake, d_logits_fake = discriminator(g_model, reuse=True)
    
    d_loss_real = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_real,
                                                labels=(tf.ones_like(d_model_real)*(1-smooth)) ))
    
    d_loss_fake = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake,
                                                labels=tf.zeros_like(d_model_fake)))

    dicriminator_loss = d_loss_real + d_loss_fake
    
    g_loss = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake, labels=tf.ones_like(d_model_fake)))

    
    return dicriminator_loss, g_loss


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_loss(model_loss)

Tests Passed

Optimization¶

Implement model_opt to create the optimization operations for the GANs. Use tf.trainable_variables to get all the trainable variables. Filter the variables with names that are in the discriminator and generator scope names. The function should return a tuple of (discriminator training operation, generator training operation).

def model_opt(d_loss, g_loss, learning_rate, beta1):
    """
    Get optimization operations
    :param d_loss: Discriminator loss Tensor
    :param g_loss: Generator loss Tensor
    :param learning_rate: Learning Rate Placeholder
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :return: A tuple of (discriminator training operation, generator training operation)
    """
    # TODO: Implement Function
    # Get weights and bias to update
    t_vars = tf.trainable_variables()
    d_vars = [var for var in t_vars if var.name.startswith('discriminator')]
    g_vars = [var for var in t_vars if var.name.startswith('generator')]

    # Optimize
    with tf.control_dependencies(tf.get_collection(tf.GraphKeys.UPDATE_OPS)):
        d_train_opt = tf.train.AdamOptimizer(learning_rate, beta1=beta1).minimize(d_loss, var_list=d_vars)
        g_train_opt = tf.train.AdamOptimizer(learning_rate, beta1=beta1).minimize(g_loss, var_list=g_vars)
    
    return d_train_opt, g_train_opt


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_opt(model_opt, tf)

Tests Passed

Neural Network Training¶

Show Output¶

Use this function to show the current output of the generator during training. It will help you determine how well the GANs is training.

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import numpy as np

def show_generator_output(sess, n_images, input_z, out_channel_dim, image_mode):
    """
    Show example output for the generator
    :param sess: TensorFlow session
    :param n_images: Number of Images to display
    :param input_z: Input Z Tensor
    :param out_channel_dim: The number of channels in the output image
    :param image_mode: The mode to use for images ("RGB" or "L")
    """
    cmap = None if image_mode == 'RGB' else 'gray'
    z_dim = input_z.get_shape().as_list()[-1]
    example_z = np.random.uniform(-1, 1, size=[n_images, z_dim])

    samples = sess.run(
        generator(input_z, out_channel_dim, False),
        feed_dict={input_z: example_z})

    images_grid = helper.images_square_grid(samples, image_mode)
    pyplot.imshow(images_grid, cmap=cmap)
    pyplot.show()

Train¶

Implement train to build and train the GANs. Use the following functions you implemented:

model_inputs(image_width, image_height, image_channels, z_dim)
model_loss(input_real, input_z, out_channel_dim)
model_opt(d_loss, g_loss, learning_rate, beta1)

Use the show_generator_output to show generator output while you train. Running show_generator_output for every batch will drastically increase training time and increase the size of the notebook. It's recommended to print the generator output every 100 batches.

def train(epoch_count, batch_size, z_dim, learning_rate, beta1, get_batches, data_shape, data_image_mode):
    """
    Train the GAN
    :param epoch_count: Number of epochs
    :param batch_size: Batch Size
    :param z_dim: Z dimension
    :param learning_rate: Learning Rate
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :param get_batches: Function to get batches
    :param data_shape: Shape of the data
    :param data_image_mode: The image mode to use for images ("RGB" or "L")
    """
    # TODO: Build Model
    print("Shape:", data_shape)
    print("Learning rate:", learning_rate)
    img_width = data_shape[1]
    img_height = data_shape[2]
    img_channels = data_shape[3]
    
    # Init GAN
    input_real, input_z, lr = model_inputs(img_width, img_height, img_channels, z_dim)
    print(input_real)
    print(input_z)
    d_loss, g_loss = model_loss(input_real, input_z, img_channels)
    d_opt, g_opt = model_opt(d_loss, g_loss, learning_rate, beta1)
    
    # Train
    steps = 0
    saver = tf.train.Saver()    
    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())
        for epoch_i in range(epoch_count):
            for batch_images in get_batches(batch_size):
                # TODO: Train Model
                steps += 1
                
                #adjust the range
                batch_images *= 2
                
                batch_z = np.random.uniform(-1,1, size=(batch_size, z_dim))
                
                _ = sess.run(d_opt, feed_dict={input_real: batch_images, input_z: batch_z, lr: learning_rate})
                _ = sess.run(g_opt, feed_dict={input_real: batch_images, input_z: batch_z, lr: learning_rate})
                
                if steps % 5 == 0:
                    d_loss_train = d_loss.eval({input_z: batch_z, input_real: batch_images})
                    g_loss_train = g_loss.eval({input_z: batch_z})
                    
                    print("Epoch {}/{}...".format(epoch_i+1, epoch_count),
                         "Discriminator Loss: {:.4f}...".format(d_loss_train),
                         "Generator Loss: {:.4f}".format(g_loss_train))
                    
                if steps % 100 == 0:
                    show_generator_output(sess, 25, input_z, img_channels, data_image_mode)

MNIST¶

Test your GANs architecture on MNIST. After 2 epochs, the GANs should be able to generate images that look like handwritten digits. Make sure the loss of the generator is lower than the loss of the discriminator or close to 0.

batch_size = 64
z_dim = 128
learning_rate = 0.0002
beta1 = 0.5

tf.reset_default_graph()
"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 3

mnist_dataset = helper.Dataset('mnist', glob(os.path.join(data_dir, 'mnist/*.jpg')))

with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, mnist_dataset.get_batches,
          mnist_dataset.shape, mnist_dataset.image_mode)

Shape: (60000, 28, 28, 1)
Learning rate: 0.0002
Tensor("input_real:0", shape=(?, 28, 28, 1), dtype=float32)
Tensor("input_z:0", shape=(?, 128), dtype=float32)
Epoch 1/3... Discriminator Loss: 4.7659... Generator Loss: 0.0188
Epoch 1/3... Discriminator Loss: 4.4947... Generator Loss: 0.0233
Epoch 1/3... Discriminator Loss: 0.4027... Generator Loss: 3.8965
Epoch 1/3... Discriminator Loss: 2.2853... Generator Loss: 0.1931
Epoch 1/3... Discriminator Loss: 0.9760... Generator Loss: 0.9237
Epoch 1/3... Discriminator Loss: 0.7341... Generator Loss: 1.2439
Epoch 1/3... Discriminator Loss: 0.4050... Generator Loss: 2.9258
Epoch 1/3... Discriminator Loss: 0.3604... Generator Loss: 7.8796
Epoch 1/3... Discriminator Loss: 0.3446... Generator Loss: 6.7756
Epoch 1/3... Discriminator Loss: 0.3490... Generator Loss: 4.7021
Epoch 1/3... Discriminator Loss: 0.3380... Generator Loss: 7.8943
Epoch 1/3... Discriminator Loss: 0.4240... Generator Loss: 2.8741
Epoch 1/3... Discriminator Loss: 0.4304... Generator Loss: 2.6131
Epoch 1/3... Discriminator Loss: 0.5813... Generator Loss: 1.7513
Epoch 1/3... Discriminator Loss: 1.1915... Generator Loss: 0.6438
Epoch 1/3... Discriminator Loss: 0.9627... Generator Loss: 1.0135
Epoch 1/3... Discriminator Loss: 0.7917... Generator Loss: 1.3625
Epoch 1/3... Discriminator Loss: 0.6223... Generator Loss: 1.5891
Epoch 1/3... Discriminator Loss: 0.4546... Generator Loss: 2.4182
Epoch 1/3... Discriminator Loss: 0.4179... Generator Loss: 2.8465

Epoch 1/3... Discriminator Loss: 0.3845... Generator Loss: 3.1943
Epoch 1/3... Discriminator Loss: 0.3818... Generator Loss: 3.3568
Epoch 1/3... Discriminator Loss: 0.3741... Generator Loss: 3.3084
Epoch 1/3... Discriminator Loss: 0.3626... Generator Loss: 3.6717
Epoch 1/3... Discriminator Loss: 0.3694... Generator Loss: 3.4024
Epoch 1/3... Discriminator Loss: 1.9536... Generator Loss: 0.3407
Epoch 1/3... Discriminator Loss: 2.5141... Generator Loss: 0.3181
Epoch 1/3... Discriminator Loss: 1.0397... Generator Loss: 1.4639
Epoch 1/3... Discriminator Loss: 2.4058... Generator Loss: 0.1648
Epoch 1/3... Discriminator Loss: 1.2593... Generator Loss: 0.8625
Epoch 1/3... Discriminator Loss: 2.3735... Generator Loss: 2.0108
Epoch 1/3... Discriminator Loss: 1.0118... Generator Loss: 1.0695
Epoch 1/3... Discriminator Loss: 1.5564... Generator Loss: 0.7025
Epoch 1/3... Discriminator Loss: 1.2587... Generator Loss: 0.6891
Epoch 1/3... Discriminator Loss: 1.1658... Generator Loss: 1.2454
Epoch 1/3... Discriminator Loss: 1.5093... Generator Loss: 0.5005
Epoch 1/3... Discriminator Loss: 1.6412... Generator Loss: 0.7673
Epoch 1/3... Discriminator Loss: 1.2932... Generator Loss: 0.7701
Epoch 1/3... Discriminator Loss: 1.3640... Generator Loss: 0.7209
Epoch 1/3... Discriminator Loss: 1.3159... Generator Loss: 0.6645

Epoch 1/3... Discriminator Loss: 1.4971... Generator Loss: 0.6613
Epoch 1/3... Discriminator Loss: 1.4134... Generator Loss: 0.6461
Epoch 1/3... Discriminator Loss: 1.5186... Generator Loss: 1.1097
Epoch 1/3... Discriminator Loss: 1.4151... Generator Loss: 0.6535
Epoch 1/3... Discriminator Loss: 1.5168... Generator Loss: 0.6121
Epoch 1/3... Discriminator Loss: 1.5880... Generator Loss: 0.5163
Epoch 1/3... Discriminator Loss: 1.5745... Generator Loss: 0.5369
Epoch 1/3... Discriminator Loss: 1.3387... Generator Loss: 0.8179
Epoch 1/3... Discriminator Loss: 1.3308... Generator Loss: 0.9531
Epoch 1/3... Discriminator Loss: 1.4325... Generator Loss: 0.6782
Epoch 1/3... Discriminator Loss: 1.4776... Generator Loss: 0.9106
Epoch 1/3... Discriminator Loss: 1.4483... Generator Loss: 0.8055
Epoch 1/3... Discriminator Loss: 1.3740... Generator Loss: 0.7570
Epoch 1/3... Discriminator Loss: 1.4544... Generator Loss: 0.9056
Epoch 1/3... Discriminator Loss: 1.4018... Generator Loss: 0.8191
Epoch 1/3... Discriminator Loss: 1.4104... Generator Loss: 0.6950
Epoch 1/3... Discriminator Loss: 1.4973... Generator Loss: 0.6547
Epoch 1/3... Discriminator Loss: 1.2122... Generator Loss: 1.0743
Epoch 1/3... Discriminator Loss: 1.5348... Generator Loss: 0.7020
Epoch 1/3... Discriminator Loss: 1.3654... Generator Loss: 0.6872

Epoch 1/3... Discriminator Loss: 1.3057... Generator Loss: 0.7758
Epoch 1/3... Discriminator Loss: 1.3704... Generator Loss: 0.8114
Epoch 1/3... Discriminator Loss: 1.4318... Generator Loss: 0.7901
Epoch 1/3... Discriminator Loss: 1.2055... Generator Loss: 0.8604
Epoch 1/3... Discriminator Loss: 1.4201... Generator Loss: 0.6619
Epoch 1/3... Discriminator Loss: 1.5538... Generator Loss: 0.5477
Epoch 1/3... Discriminator Loss: 1.3319... Generator Loss: 0.7572
Epoch 1/3... Discriminator Loss: 1.3438... Generator Loss: 0.7505
Epoch 1/3... Discriminator Loss: 1.5113... Generator Loss: 0.8267
Epoch 1/3... Discriminator Loss: 1.3558... Generator Loss: 0.9177
Epoch 1/3... Discriminator Loss: 1.3990... Generator Loss: 0.8248
Epoch 1/3... Discriminator Loss: 1.3395... Generator Loss: 0.8285
Epoch 1/3... Discriminator Loss: 1.2768... Generator Loss: 0.7980
Epoch 1/3... Discriminator Loss: 1.2935... Generator Loss: 0.7164
Epoch 1/3... Discriminator Loss: 1.3798... Generator Loss: 0.6853
Epoch 1/3... Discriminator Loss: 1.3399... Generator Loss: 0.9032
Epoch 1/3... Discriminator Loss: 1.2874... Generator Loss: 1.0035
Epoch 1/3... Discriminator Loss: 1.2371... Generator Loss: 0.8140
Epoch 1/3... Discriminator Loss: 1.4151... Generator Loss: 0.8574
Epoch 1/3... Discriminator Loss: 1.2661... Generator Loss: 1.0870

Epoch 1/3... Discriminator Loss: 1.3201... Generator Loss: 0.8242
Epoch 1/3... Discriminator Loss: 1.3076... Generator Loss: 0.8090
Epoch 1/3... Discriminator Loss: 1.2097... Generator Loss: 0.8756
Epoch 1/3... Discriminator Loss: 1.3074... Generator Loss: 0.7320
Epoch 1/3... Discriminator Loss: 1.2661... Generator Loss: 0.7547
Epoch 1/3... Discriminator Loss: 1.2878... Generator Loss: 0.7277
Epoch 1/3... Discriminator Loss: 1.1698... Generator Loss: 0.9452
Epoch 1/3... Discriminator Loss: 1.1823... Generator Loss: 0.9228
Epoch 1/3... Discriminator Loss: 1.2609... Generator Loss: 0.9783
Epoch 1/3... Discriminator Loss: 1.2676... Generator Loss: 0.7598
Epoch 1/3... Discriminator Loss: 1.1903... Generator Loss: 1.0124
Epoch 1/3... Discriminator Loss: 1.2453... Generator Loss: 0.7990
Epoch 1/3... Discriminator Loss: 1.2567... Generator Loss: 1.1764
Epoch 1/3... Discriminator Loss: 1.2333... Generator Loss: 0.9197
Epoch 1/3... Discriminator Loss: 1.1776... Generator Loss: 0.9776
Epoch 1/3... Discriminator Loss: 1.2410... Generator Loss: 0.9002
Epoch 1/3... Discriminator Loss: 1.2158... Generator Loss: 1.1097
Epoch 1/3... Discriminator Loss: 1.2190... Generator Loss: 0.7352
Epoch 1/3... Discriminator Loss: 1.1785... Generator Loss: 0.9259
Epoch 1/3... Discriminator Loss: 1.2205... Generator Loss: 0.8506

Epoch 1/3... Discriminator Loss: 1.3289... Generator Loss: 0.7298
Epoch 1/3... Discriminator Loss: 1.1101... Generator Loss: 0.8957
Epoch 1/3... Discriminator Loss: 1.1442... Generator Loss: 1.0195
Epoch 1/3... Discriminator Loss: 1.1889... Generator Loss: 1.0632
Epoch 1/3... Discriminator Loss: 1.2669... Generator Loss: 1.0061
Epoch 1/3... Discriminator Loss: 1.2272... Generator Loss: 0.9109
Epoch 1/3... Discriminator Loss: 1.2307... Generator Loss: 0.7594
Epoch 1/3... Discriminator Loss: 1.2083... Generator Loss: 0.8181
Epoch 1/3... Discriminator Loss: 1.1896... Generator Loss: 1.0182
Epoch 1/3... Discriminator Loss: 1.1720... Generator Loss: 1.2383
Epoch 1/3... Discriminator Loss: 1.1948... Generator Loss: 0.9520
Epoch 1/3... Discriminator Loss: 1.1115... Generator Loss: 1.0988
Epoch 1/3... Discriminator Loss: 1.1807... Generator Loss: 0.8920
Epoch 1/3... Discriminator Loss: 1.1639... Generator Loss: 0.9698
Epoch 1/3... Discriminator Loss: 1.2131... Generator Loss: 0.8371
Epoch 1/3... Discriminator Loss: 1.1970... Generator Loss: 1.0277
Epoch 1/3... Discriminator Loss: 1.1358... Generator Loss: 1.0222
Epoch 1/3... Discriminator Loss: 1.1893... Generator Loss: 0.8348
Epoch 1/3... Discriminator Loss: 1.1762... Generator Loss: 1.0332
Epoch 1/3... Discriminator Loss: 1.1677... Generator Loss: 0.9052

Epoch 1/3... Discriminator Loss: 1.1882... Generator Loss: 0.8828
Epoch 1/3... Discriminator Loss: 1.1286... Generator Loss: 0.9676
Epoch 1/3... Discriminator Loss: 1.1700... Generator Loss: 0.9910
Epoch 1/3... Discriminator Loss: 1.2649... Generator Loss: 0.7454
Epoch 1/3... Discriminator Loss: 1.2024... Generator Loss: 0.8008
Epoch 1/3... Discriminator Loss: 1.1743... Generator Loss: 0.8748
Epoch 1/3... Discriminator Loss: 1.1644... Generator Loss: 1.1911
Epoch 1/3... Discriminator Loss: 1.2225... Generator Loss: 1.2383
Epoch 1/3... Discriminator Loss: 1.1869... Generator Loss: 0.9570
Epoch 1/3... Discriminator Loss: 1.1342... Generator Loss: 1.0373
Epoch 1/3... Discriminator Loss: 1.2002... Generator Loss: 0.8206
Epoch 1/3... Discriminator Loss: 1.1731... Generator Loss: 0.8212
Epoch 1/3... Discriminator Loss: 1.1185... Generator Loss: 1.0883
Epoch 1/3... Discriminator Loss: 1.1714... Generator Loss: 1.0868
Epoch 1/3... Discriminator Loss: 1.0812... Generator Loss: 1.0013
Epoch 1/3... Discriminator Loss: 1.1733... Generator Loss: 0.9287
Epoch 1/3... Discriminator Loss: 1.1429... Generator Loss: 1.0974
Epoch 1/3... Discriminator Loss: 1.1146... Generator Loss: 0.9060
Epoch 1/3... Discriminator Loss: 1.1411... Generator Loss: 0.9080
Epoch 1/3... Discriminator Loss: 1.1342... Generator Loss: 0.9064

Epoch 1/3... Discriminator Loss: 1.1296... Generator Loss: 1.0497
Epoch 1/3... Discriminator Loss: 1.1602... Generator Loss: 0.9260
Epoch 1/3... Discriminator Loss: 1.0585... Generator Loss: 1.1040
Epoch 1/3... Discriminator Loss: 1.0911... Generator Loss: 1.0577
Epoch 1/3... Discriminator Loss: 1.1109... Generator Loss: 1.0677
Epoch 1/3... Discriminator Loss: 1.1026... Generator Loss: 0.9726
Epoch 1/3... Discriminator Loss: 1.1028... Generator Loss: 0.8835
Epoch 1/3... Discriminator Loss: 1.1402... Generator Loss: 0.8818
Epoch 1/3... Discriminator Loss: 1.0853... Generator Loss: 0.9517
Epoch 1/3... Discriminator Loss: 1.1071... Generator Loss: 1.0022
Epoch 1/3... Discriminator Loss: 1.0770... Generator Loss: 1.1828
Epoch 1/3... Discriminator Loss: 1.1294... Generator Loss: 0.9742
Epoch 1/3... Discriminator Loss: 1.1627... Generator Loss: 0.8631
Epoch 1/3... Discriminator Loss: 1.0991... Generator Loss: 1.1651
Epoch 1/3... Discriminator Loss: 1.1019... Generator Loss: 1.1490
Epoch 1/3... Discriminator Loss: 1.1192... Generator Loss: 1.1969
Epoch 1/3... Discriminator Loss: 1.1546... Generator Loss: 0.8795
Epoch 1/3... Discriminator Loss: 1.2257... Generator Loss: 0.8373
Epoch 1/3... Discriminator Loss: 1.1244... Generator Loss: 0.8424
Epoch 1/3... Discriminator Loss: 1.1455... Generator Loss: 1.0839

Epoch 1/3... Discriminator Loss: 1.1279... Generator Loss: 1.0353
Epoch 1/3... Discriminator Loss: 1.1724... Generator Loss: 0.7799
Epoch 1/3... Discriminator Loss: 1.1493... Generator Loss: 1.0101
Epoch 1/3... Discriminator Loss: 1.0917... Generator Loss: 0.8688
Epoch 1/3... Discriminator Loss: 1.0900... Generator Loss: 0.9639
Epoch 1/3... Discriminator Loss: 1.1204... Generator Loss: 1.0307
Epoch 1/3... Discriminator Loss: 1.1418... Generator Loss: 0.8828
Epoch 1/3... Discriminator Loss: 1.1603... Generator Loss: 0.9034
Epoch 1/3... Discriminator Loss: 1.0990... Generator Loss: 1.0161
Epoch 1/3... Discriminator Loss: 1.1320... Generator Loss: 1.0592
Epoch 1/3... Discriminator Loss: 1.1133... Generator Loss: 1.1379
Epoch 1/3... Discriminator Loss: 1.1179... Generator Loss: 1.0798
Epoch 1/3... Discriminator Loss: 1.1450... Generator Loss: 1.1443
Epoch 1/3... Discriminator Loss: 1.1038... Generator Loss: 1.4317
Epoch 1/3... Discriminator Loss: 1.1311... Generator Loss: 0.9563
Epoch 1/3... Discriminator Loss: 1.1032... Generator Loss: 0.9542
Epoch 1/3... Discriminator Loss: 1.1258... Generator Loss: 1.0131
Epoch 1/3... Discriminator Loss: 1.1217... Generator Loss: 0.9537
Epoch 1/3... Discriminator Loss: 1.0763... Generator Loss: 0.9944
Epoch 1/3... Discriminator Loss: 1.1610... Generator Loss: 0.8280

Epoch 1/3... Discriminator Loss: 1.1216... Generator Loss: 0.9797
Epoch 1/3... Discriminator Loss: 1.0496... Generator Loss: 1.2591
Epoch 1/3... Discriminator Loss: 1.1747... Generator Loss: 0.7089
Epoch 1/3... Discriminator Loss: 1.1028... Generator Loss: 1.0079
Epoch 1/3... Discriminator Loss: 1.1474... Generator Loss: 0.9943
Epoch 1/3... Discriminator Loss: 1.0923... Generator Loss: 1.1973
Epoch 1/3... Discriminator Loss: 1.1149... Generator Loss: 1.3261
Epoch 2/3... Discriminator Loss: 1.2256... Generator Loss: 0.8183
Epoch 2/3... Discriminator Loss: 1.2178... Generator Loss: 0.8673
Epoch 2/3... Discriminator Loss: 1.1570... Generator Loss: 1.0371
Epoch 2/3... Discriminator Loss: 1.1241... Generator Loss: 0.9367
Epoch 2/3... Discriminator Loss: 1.1736... Generator Loss: 1.1655
Epoch 2/3... Discriminator Loss: 1.1929... Generator Loss: 0.9998
Epoch 2/3... Discriminator Loss: 1.1087... Generator Loss: 0.9360
Epoch 2/3... Discriminator Loss: 1.1582... Generator Loss: 0.9288
Epoch 2/3... Discriminator Loss: 1.2469... Generator Loss: 1.1430
Epoch 2/3... Discriminator Loss: 1.1700... Generator Loss: 0.9061
Epoch 2/3... Discriminator Loss: 1.1930... Generator Loss: 0.9375
Epoch 2/3... Discriminator Loss: 1.1256... Generator Loss: 1.0324
Epoch 2/3... Discriminator Loss: 1.2146... Generator Loss: 1.1126

CelebA¶

Run your GANs on CelebA. It will take around 20 minutes on the average GPU to run one epoch. You can run the whole epoch or stop when it starts to generate realistic faces.

batch_size = 128
z_dim = 128
learning_rate = 0.0002
beta1 = 0.5


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 6

celeba_dataset = helper.Dataset('celeba', glob(os.path.join(data_dir, 'img_align_celeba/*.jpg')))
with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, celeba_dataset.get_batches,
          celeba_dataset.shape, celeba_dataset.image_mode)

Shape: (202599, 28, 28, 3)
Learning rate: 0.0002
Tensor("input_real:0", shape=(?, 28, 28, 3), dtype=float32)
Tensor("input_z:0", shape=(?, 128), dtype=float32)
Epoch 1/6... Discriminator Loss: 1.1003... Generator Loss: 0.7900
Epoch 1/6... Discriminator Loss: 0.9388... Generator Loss: 1.0178
Epoch 1/6... Discriminator Loss: 1.0697... Generator Loss: 0.8986
Epoch 1/6... Discriminator Loss: 0.4054... Generator Loss: 4.0422
Epoch 1/6... Discriminator Loss: 2.2092... Generator Loss: 0.2261
Epoch 1/6... Discriminator Loss: 0.4554... Generator Loss: 9.2646
Epoch 1/6... Discriminator Loss: 0.5326... Generator Loss: 2.9882
Epoch 1/6... Discriminator Loss: 8.3735... Generator Loss: 0.0004
Epoch 1/6... Discriminator Loss: 2.6870... Generator Loss: 0.1292
Epoch 1/6... Discriminator Loss: 0.4101... Generator Loss: 3.5363
Epoch 1/6... Discriminator Loss: 0.4695... Generator Loss: 2.6124
Epoch 1/6... Discriminator Loss: 0.7698... Generator Loss: 1.2171
Epoch 1/6... Discriminator Loss: 0.5244... Generator Loss: 2.0774
Epoch 1/6... Discriminator Loss: 0.5302... Generator Loss: 2.0558
Epoch 1/6... Discriminator Loss: 0.4310... Generator Loss: 4.2416
Epoch 1/6... Discriminator Loss: 0.4480... Generator Loss: 2.8505
Epoch 1/6... Discriminator Loss: 0.4682... Generator Loss: 2.4485
Epoch 1/6... Discriminator Loss: 0.6388... Generator Loss: 1.7460
Epoch 1/6... Discriminator Loss: 0.5873... Generator Loss: 1.9586
Epoch 1/6... Discriminator Loss: 0.6549... Generator Loss: 1.5292

Epoch 1/6... Discriminator Loss: 0.6228... Generator Loss: 1.6564
Epoch 1/6... Discriminator Loss: 0.6488... Generator Loss: 1.6577
Epoch 1/6... Discriminator Loss: 0.6279... Generator Loss: 1.7084
Epoch 1/6... Discriminator Loss: 0.6578... Generator Loss: 1.6229
Epoch 1/6... Discriminator Loss: 0.7206... Generator Loss: 1.4052
Epoch 1/6... Discriminator Loss: 0.7952... Generator Loss: 1.2978
Epoch 1/6... Discriminator Loss: 0.7901... Generator Loss: 1.5375
Epoch 1/6... Discriminator Loss: 1.5078... Generator Loss: 0.9089
Epoch 1/6... Discriminator Loss: 1.0937... Generator Loss: 1.1602
Epoch 1/6... Discriminator Loss: 1.3590... Generator Loss: 1.9311
Epoch 1/6... Discriminator Loss: 1.5709... Generator Loss: 0.5012
Epoch 1/6... Discriminator Loss: 1.3398... Generator Loss: 1.3913
Epoch 1/6... Discriminator Loss: 1.4425... Generator Loss: 0.5507
Epoch 1/6... Discriminator Loss: 1.5579... Generator Loss: 1.7484
Epoch 1/6... Discriminator Loss: 1.1943... Generator Loss: 0.9195
Epoch 1/6... Discriminator Loss: 1.2512... Generator Loss: 1.4152
Epoch 1/6... Discriminator Loss: 1.1737... Generator Loss: 0.8575
Epoch 1/6... Discriminator Loss: 1.2454... Generator Loss: 0.7657
Epoch 1/6... Discriminator Loss: 1.4312... Generator Loss: 1.6328
Epoch 1/6... Discriminator Loss: 1.2688... Generator Loss: 0.7584

Epoch 1/6... Discriminator Loss: 1.2402... Generator Loss: 0.8628
Epoch 1/6... Discriminator Loss: 1.4672... Generator Loss: 1.7479
Epoch 1/6... Discriminator Loss: 1.2976... Generator Loss: 0.7367
Epoch 1/6... Discriminator Loss: 1.4915... Generator Loss: 1.4929
Epoch 1/6... Discriminator Loss: 1.2463... Generator Loss: 0.8187
Epoch 1/6... Discriminator Loss: 1.2329... Generator Loss: 0.9029
Epoch 1/6... Discriminator Loss: 1.0811... Generator Loss: 1.1491
Epoch 1/6... Discriminator Loss: 1.1230... Generator Loss: 0.9647
Epoch 1/6... Discriminator Loss: 0.9726... Generator Loss: 1.4035
Epoch 1/6... Discriminator Loss: 1.4019... Generator Loss: 0.5640
Epoch 1/6... Discriminator Loss: 1.2713... Generator Loss: 1.1736
Epoch 1/6... Discriminator Loss: 1.3079... Generator Loss: 0.9922
Epoch 1/6... Discriminator Loss: 1.4705... Generator Loss: 0.4936
Epoch 1/6... Discriminator Loss: 1.3726... Generator Loss: 1.3309
Epoch 1/6... Discriminator Loss: 1.3294... Generator Loss: 0.8126
Epoch 1/6... Discriminator Loss: 1.5577... Generator Loss: 0.4521
Epoch 1/6... Discriminator Loss: 1.4643... Generator Loss: 0.7393
Epoch 1/6... Discriminator Loss: 1.4451... Generator Loss: 1.0363
Epoch 1/6... Discriminator Loss: 1.5021... Generator Loss: 0.8223
Epoch 1/6... Discriminator Loss: 1.3444... Generator Loss: 0.8568

Epoch 1/6... Discriminator Loss: 1.5603... Generator Loss: 0.5449
Epoch 1/6... Discriminator Loss: 1.3245... Generator Loss: 0.9294
Epoch 1/6... Discriminator Loss: 1.5615... Generator Loss: 0.4418
Epoch 1/6... Discriminator Loss: 1.4266... Generator Loss: 0.7775
Epoch 1/6... Discriminator Loss: 1.2559... Generator Loss: 0.8599
Epoch 1/6... Discriminator Loss: 1.4377... Generator Loss: 0.8386
Epoch 1/6... Discriminator Loss: 1.4460... Generator Loss: 0.8100
Epoch 1/6... Discriminator Loss: 1.4478... Generator Loss: 0.6229
Epoch 1/6... Discriminator Loss: 1.7227... Generator Loss: 1.3015
Epoch 1/6... Discriminator Loss: 1.3315... Generator Loss: 1.2494
Epoch 1/6... Discriminator Loss: 1.5011... Generator Loss: 0.5450
Epoch 1/6... Discriminator Loss: 1.4940... Generator Loss: 1.2344
Epoch 1/6... Discriminator Loss: 1.4517... Generator Loss: 0.6968
Epoch 1/6... Discriminator Loss: 1.4099... Generator Loss: 0.5734
Epoch 1/6... Discriminator Loss: 1.4349... Generator Loss: 0.8062
Epoch 1/6... Discriminator Loss: 1.3514... Generator Loss: 0.6699
Epoch 1/6... Discriminator Loss: 1.2948... Generator Loss: 0.7694
Epoch 1/6... Discriminator Loss: 1.3923... Generator Loss: 1.0065
Epoch 1/6... Discriminator Loss: 1.6532... Generator Loss: 0.6962
Epoch 1/6... Discriminator Loss: 1.4419... Generator Loss: 0.7042

Epoch 1/6... Discriminator Loss: 1.3159... Generator Loss: 0.6398
Epoch 1/6... Discriminator Loss: 1.4042... Generator Loss: 0.7607
Epoch 1/6... Discriminator Loss: 1.4752... Generator Loss: 0.6699
Epoch 1/6... Discriminator Loss: 1.6546... Generator Loss: 0.8809
Epoch 1/6... Discriminator Loss: 1.3286... Generator Loss: 0.6392
Epoch 1/6... Discriminator Loss: 1.4147... Generator Loss: 0.7813
Epoch 1/6... Discriminator Loss: 1.4610... Generator Loss: 0.6789
Epoch 1/6... Discriminator Loss: 1.3097... Generator Loss: 1.0637
Epoch 1/6... Discriminator Loss: 1.5795... Generator Loss: 0.4498
Epoch 1/6... Discriminator Loss: 1.2107... Generator Loss: 0.9410
Epoch 1/6... Discriminator Loss: 1.3599... Generator Loss: 0.8327
Epoch 1/6... Discriminator Loss: 1.4190... Generator Loss: 0.5925
Epoch 1/6... Discriminator Loss: 1.3150... Generator Loss: 0.9903
Epoch 1/6... Discriminator Loss: 1.3896... Generator Loss: 0.7476
Epoch 1/6... Discriminator Loss: 1.5454... Generator Loss: 0.4985
Epoch 1/6... Discriminator Loss: 1.3715... Generator Loss: 0.8608
Epoch 1/6... Discriminator Loss: 1.5079... Generator Loss: 0.7416
Epoch 1/6... Discriminator Loss: 1.4032... Generator Loss: 0.7423
Epoch 1/6... Discriminator Loss: 1.5024... Generator Loss: 0.7050
Epoch 1/6... Discriminator Loss: 1.6315... Generator Loss: 0.6634

Epoch 1/6... Discriminator Loss: 1.3359... Generator Loss: 0.8126
Epoch 1/6... Discriminator Loss: 1.4373... Generator Loss: 0.6911
Epoch 1/6... Discriminator Loss: 1.5393... Generator Loss: 0.7821
Epoch 1/6... Discriminator Loss: 1.4081... Generator Loss: 0.9506
Epoch 1/6... Discriminator Loss: 1.4257... Generator Loss: 0.8916
Epoch 1/6... Discriminator Loss: 1.3607... Generator Loss: 0.7146
Epoch 1/6... Discriminator Loss: 1.3094... Generator Loss: 1.0637
Epoch 1/6... Discriminator Loss: 1.4814... Generator Loss: 0.6392
Epoch 1/6... Discriminator Loss: 1.2151... Generator Loss: 1.0622
Epoch 1/6... Discriminator Loss: 1.5624... Generator Loss: 0.5444
Epoch 1/6... Discriminator Loss: 1.4897... Generator Loss: 0.6766
Epoch 1/6... Discriminator Loss: 1.3237... Generator Loss: 0.9787
Epoch 1/6... Discriminator Loss: 1.2419... Generator Loss: 0.8292
Epoch 1/6... Discriminator Loss: 1.3748... Generator Loss: 0.8248
Epoch 1/6... Discriminator Loss: 1.3481... Generator Loss: 0.8042
Epoch 1/6... Discriminator Loss: 1.6278... Generator Loss: 0.6131
Epoch 1/6... Discriminator Loss: 1.3889... Generator Loss: 0.8520
Epoch 1/6... Discriminator Loss: 1.4692... Generator Loss: 0.6633
Epoch 1/6... Discriminator Loss: 1.3592... Generator Loss: 0.7593
Epoch 1/6... Discriminator Loss: 1.4084... Generator Loss: 0.6393

Epoch 1/6... Discriminator Loss: 1.3686... Generator Loss: 0.7302
Epoch 1/6... Discriminator Loss: 1.3760... Generator Loss: 0.7518
Epoch 1/6... Discriminator Loss: 1.3532... Generator Loss: 0.7769
Epoch 1/6... Discriminator Loss: 1.2486... Generator Loss: 0.8263
Epoch 1/6... Discriminator Loss: 1.0528... Generator Loss: 0.8035
Epoch 1/6... Discriminator Loss: 0.7016... Generator Loss: 1.6012
Epoch 1/6... Discriminator Loss: 1.3276... Generator Loss: 0.5586
Epoch 1/6... Discriminator Loss: 1.7418... Generator Loss: 2.0584
Epoch 1/6... Discriminator Loss: 0.6314... Generator Loss: 2.7759
Epoch 1/6... Discriminator Loss: 0.9195... Generator Loss: 0.9776
Epoch 1/6... Discriminator Loss: 0.4072... Generator Loss: 3.2936
Epoch 1/6... Discriminator Loss: 0.4974... Generator Loss: 4.2853
Epoch 1/6... Discriminator Loss: 0.7077... Generator Loss: 1.3768
Epoch 1/6... Discriminator Loss: 0.4488... Generator Loss: 4.6728
Epoch 1/6... Discriminator Loss: 0.5220... Generator Loss: 2.3253
Epoch 1/6... Discriminator Loss: 0.5428... Generator Loss: 3.1851
Epoch 1/6... Discriminator Loss: 0.9844... Generator Loss: 0.8135
Epoch 1/6... Discriminator Loss: 0.6228... Generator Loss: 4.9306
Epoch 1/6... Discriminator Loss: 0.4354... Generator Loss: 2.6983
Epoch 1/6... Discriminator Loss: 0.4123... Generator Loss: 3.2821

Epoch 1/6... Discriminator Loss: 0.9224... Generator Loss: 0.9420
Epoch 1/6... Discriminator Loss: 0.9703... Generator Loss: 2.2632
Epoch 1/6... Discriminator Loss: 0.4086... Generator Loss: 3.7993
Epoch 1/6... Discriminator Loss: 0.7560... Generator Loss: 3.1472
Epoch 1/6... Discriminator Loss: 0.8233... Generator Loss: 1.0915
Epoch 1/6... Discriminator Loss: 0.4639... Generator Loss: 2.5997
Epoch 1/6... Discriminator Loss: 0.5200... Generator Loss: 2.0955
Epoch 1/6... Discriminator Loss: 1.1042... Generator Loss: 6.4659
Epoch 1/6... Discriminator Loss: 0.6887... Generator Loss: 1.6045
Epoch 1/6... Discriminator Loss: 1.3585... Generator Loss: 0.5068
Epoch 1/6... Discriminator Loss: 0.6027... Generator Loss: 1.5869
Epoch 1/6... Discriminator Loss: 0.8393... Generator Loss: 1.0458
Epoch 1/6... Discriminator Loss: 0.7268... Generator Loss: 1.5335
Epoch 1/6... Discriminator Loss: 1.0383... Generator Loss: 0.7718
Epoch 1/6... Discriminator Loss: 0.6026... Generator Loss: 1.6947
Epoch 1/6... Discriminator Loss: 0.6344... Generator Loss: 1.9500
Epoch 1/6... Discriminator Loss: 0.4742... Generator Loss: 2.6126
Epoch 1/6... Discriminator Loss: 1.1857... Generator Loss: 0.6247
Epoch 1/6... Discriminator Loss: 1.0604... Generator Loss: 3.5428
Epoch 1/6... Discriminator Loss: 0.6168... Generator Loss: 2.6216

Epoch 1/6... Discriminator Loss: 0.5226... Generator Loss: 2.2837
Epoch 1/6... Discriminator Loss: 0.5525... Generator Loss: 2.1974
Epoch 1/6... Discriminator Loss: 0.4791... Generator Loss: 2.3664
Epoch 1/6... Discriminator Loss: 1.0777... Generator Loss: 0.7373
Epoch 1/6... Discriminator Loss: 0.5721... Generator Loss: 1.8731
Epoch 1/6... Discriminator Loss: 0.5973... Generator Loss: 2.8284
Epoch 1/6... Discriminator Loss: 1.0219... Generator Loss: 0.8370
Epoch 1/6... Discriminator Loss: 1.1135... Generator Loss: 0.8488
Epoch 1/6... Discriminator Loss: 0.6244... Generator Loss: 1.7005
Epoch 1/6... Discriminator Loss: 0.4794... Generator Loss: 2.6463
Epoch 1/6... Discriminator Loss: 0.4835... Generator Loss: 2.8017
Epoch 1/6... Discriminator Loss: 1.1483... Generator Loss: 0.6650
Epoch 1/6... Discriminator Loss: 0.6218... Generator Loss: 2.4377
Epoch 1/6... Discriminator Loss: 0.5100... Generator Loss: 2.3115
Epoch 1/6... Discriminator Loss: 0.5584... Generator Loss: 1.8837
Epoch 1/6... Discriminator Loss: 1.4877... Generator Loss: 0.4241
Epoch 1/6... Discriminator Loss: 0.8334... Generator Loss: 1.1665
Epoch 1/6... Discriminator Loss: 1.1516... Generator Loss: 1.7778
Epoch 1/6... Discriminator Loss: 1.0683... Generator Loss: 0.7248
Epoch 1/6... Discriminator Loss: 0.5749... Generator Loss: 2.0121

Epoch 1/6... Discriminator Loss: 0.5813... Generator Loss: 2.2485
Epoch 1/6... Discriminator Loss: 0.7760... Generator Loss: 2.1131
Epoch 1/6... Discriminator Loss: 1.1863... Generator Loss: 3.7799
Epoch 1/6... Discriminator Loss: 0.6219... Generator Loss: 1.6442
Epoch 1/6... Discriminator Loss: 0.5887... Generator Loss: 1.9233
Epoch 1/6... Discriminator Loss: 0.5437... Generator Loss: 2.3162
Epoch 1/6... Discriminator Loss: 2.2521... Generator Loss: 0.1939
Epoch 1/6... Discriminator Loss: 0.8708... Generator Loss: 1.2080
Epoch 1/6... Discriminator Loss: 0.9199... Generator Loss: 1.4285
Epoch 1/6... Discriminator Loss: 0.6582... Generator Loss: 1.5554
Epoch 1/6... Discriminator Loss: 0.7089... Generator Loss: 1.4684
Epoch 1/6... Discriminator Loss: 0.5779... Generator Loss: 3.4041
Epoch 1/6... Discriminator Loss: 0.6107... Generator Loss: 1.9785
Epoch 1/6... Discriminator Loss: 0.5126... Generator Loss: 2.5647
Epoch 1/6... Discriminator Loss: 0.5255... Generator Loss: 2.0112
Epoch 1/6... Discriminator Loss: 0.6214... Generator Loss: 2.0966
Epoch 1/6... Discriminator Loss: 0.5780... Generator Loss: 2.2147
Epoch 1/6... Discriminator Loss: 1.2178... Generator Loss: 1.7469
Epoch 1/6... Discriminator Loss: 0.8945... Generator Loss: 3.1274
Epoch 1/6... Discriminator Loss: 0.6221... Generator Loss: 2.0119

Submitting This Project¶

When submitting this project, make sure to run all the cells before saving the notebook. Save the notebook file as "dlnd_face_generation.ipynb" and save it as a HTML file under "File" -> "Download as". Include the "helper.py" and "problem_unittests.py" files in your submission.